Annu. Rev. Physiol. 2000. 62:135?55
Copyright q by Annual Reviews. All rights reserved
0066?4278/00/0315?0135$12.00 135
THE EVOLUTIONARY PHYSIOLOGY OF ANIMAL
FLIGHT: Paleobiological and Present Perspectives
Robert Dudley
Section of Integrative Biology, University of Texas, Austin, Texas, 78712 and
Smithsonian Tropical Research Institute, P.O. Box 2072, Balboa, Republic of Panama;
e-mail: r_dudley@utxvms.cc.utexas.edu
Key Words evolution, gigantism, hyperoxia, insect, oxygen
Abstract Recent geophysical analyses suggest the presence of a late Paleozoic
oxygen pulse beginning in the late Devonian and continuing through to the late Car-
boniferous. During this period, plant terrestrialization and global carbon deposition
resulted in a dramatic increase in atmospheric oxygen levels, ultimately yielding con-
centrations potentially as high as 35% relative to the contemporary value of 21%.
Such hyperoxia of the late Paleozoic atmosphere may have physiologically facilitated
the initial evolution of insect flight metabolism. Widespread gigantism in late Paleo-
zoic insects and other arthropods is also consistent with enhanced oxygen flux within
diffusion-limited tracheal systems. Because total atmospheric pressure increases with
increased oxygen partial pressure, concurrently hyperdense conditions would have
augmented aerodynamic force production in early forms of flying insects. By the late
Permian, evolution of decompositional microbial and fungal communities, together
with disequilibrium in rates of carbon deposition, gradually reduced oxygen concen-
trations to values possibly as low as 15%. The disappearance of giant insects by the
end of the Permian is consistent with extinction of these taxa for reasons of asphyx-
iation on a geological time scale. As with winged insects, the multiple historical
origins of vertebrate flight in the late Jurassic and Cretaceous correlate temporally
with periods of elevated atmospheric oxygen. Much discussion of flight performance
in Archaeopteryx assumes a contemporary atmospheric composition. Elevated oxygen
levels in the mid- to late Mesozoic would, however, have facilitated aerodynamic
force production and enhanced muscle power output for ancestral birds, as well as
for precursors to bats and pterosaurs.
INTRODUCTION
Today?s oxidizing atmosphere, with an oxygen concentration of about 21%,
derives in part from the metabolic activity of photosynthetic organisms. Starting
with the anoxic conditions and reducing atmosphere of the Archean and early
Proterozoic eons, evolution of cyanobacteria, sulfide-oxidizing bacteria, and algae
contributed to buildup of oxygen partial pressures by the beginning of the Cam-
136 DUDLEY
brian (570 MYa) to values likely comparable to those of the present day (1?6).
Increased atmospheric oxygen, in turn, may have been an important factor under-
lying radiations of metazoan taxa in the late Proterozoic and at the base of the
Cambrian (7?14). The paleobiological significance of Proterozoic oxygen levels
for animal evolution in aquatic habitats has long been recognized, as has been
the role of gradual oxygen buildup for subsequent evolution of plant and animal
terrestrial communities (6, 8, 9). Less well-studied, however, are potential fluc-
tuations in atmospheric constituents through the course of the Phanerozoic.
In particular, variation in oxygen partial pressure and concomitant variation in
air density potentially influence the physiology and biomechanics, respectively,
of animal flight performance. Geophysical analyses suggest the presence of a late
Paleozoic oxygen pulse beginning in the late Devonian and continuing through
to the late Carboniferous. Oxygen concentrations ranged possibly as high as 35%
during the initial period of insect flight evolution (late Devonian or early Car-
boniferous). Atmospheric hyperoxia may thus have physiologically facilitated the
initial evolution of insect flight metabolism. Because total atmospheric pressure
increases with an increased oxygen partial pressure, concurrently hyperdense con-
ditions would have augmented aerodynamic force production in ancestral winged
insects. The multiple historical origins of vertebrate flight similarly appear to
correlate temporally with periods of hyperoxia. In particular, elevated oxygen
levels in the mid- to late Mesozoic may have facilitated aerodynamic force pro-
duction and enhanced muscle metabolism for ancestral birds, as well as for pre-
cursors to bats and pterosaurs. Because of such implications for the evolution of
animal flight performance, discussion of the mechanisms underlying substantial
variation in composition of the earth?s atmosphere is warranted.
PHANEROZOIC VARIATION IN THE
EARTH?S ATMOSPHERE
The physical properties of the earth?s atmosphere are determined almost exclu-
sively by two major constituents, namely diatomic oxygen and diatomic nitrogen.
Nitrogen content of the atmosphere is thought to have been approximately con-
stant through geological time (15?17), although some variation in this constituent
may derive through biotic linkages with the atmospheric pool (CD Nevison,
unpublished data). Various models for oxygen content of the atmosphere in Phan-
erozoic times have suggested major fluctuations that are driven biotically (6, 8,
13, 16, 19, 20). In particular, geochemical modeling (21, 22) suggests a late
Paleozoic disequilibrium between the rate of carbon fixation by terrestrial plants
and the rate at which this fixed material was decomposed and recycled. Through
the mid- to end-Devonian (390?362 MYa), this disequilibrium led to a moderate
increase in oxygen concentration from 18 to 20%, followed by a more substantial
OXYGEN AND ANIMAL FLIGHT EVOLUTION 137
Figure 1 Estimate of atmospheric oxygen concentrations through the Phanerozoic (21).
PAL: present atmospheric level (20.95%).
rise by the end of the Carboniferous (290 MYa) to remarkable values as high as
35% (see Figure 1). Thereafter, an equally pronounced decline resulted in end-
Permian and early Triassic values as low as 15%. Relative to the present atmo-
spheric level of 21% (PAL), such hyperoxic and subsequently hypoxic conditions
must have had major biological effects during the Paleozoic (23? 26). Moreover,
atmospheric hyperoxia may also have characterized portions of the Mesozoic.
According to the aforementioned geochemical model (21), a secondary rise in
oxygen levels characterized the the mid-Jurassic and the Cretaceous (170?65
MYa), with oxygen concentration increasing to 25 to 30% and then declining to
PAL only by the end-Tertiary (Figure 1). Such postulated changes in oxygen
concentration are at sharp odds with any uniformitarian perspective of atmo-
spheric composition.
The various assumptions underlying reconstruction of ancient atmospheric
composition obviously require examination. Current models for Phanerozoic oxy-
gen and carbon dioxide concentrations (21, 27?29) are based on estimates of the
exchange rates of fixed and reduced carbon among atmospheric, oceanic, and
sedimentary reservoirs. Variation in rates of carbon exchange is driven primarily
by the effects of plant terrestrialization, variable rates of organic carbon deposition
and decomposition, and changes in continental weathering (20, 22, 30?32). An
additional feedback loop potentially influencing atmospheric oxygen levels
involves marine phosphorus (33, 34). Because the relative importance of these
different mechanisms can be, in many cases, only broadly constrained, consid-
erable uncertainty must be associated with such efforts that model global geo-
chemical cycles. Of particular relevance to the late Paleozoic atmosphere,
138 DUDLEY
however, are the changes necessarily associated with the global phenomenon of
terrestrialization by plants. During this process, photosynthetic production of oxy-
gen became decoupled from breakdown of fixed carbon. An ensuing decompo-
sitional bottleneck, due apparently to the absence of biotic processes leading to
rapid breakdown, resulted in extensive deposition of coal and other carbonates.
Subsequent evolution by the mid-Permian of microbial decomposers, in particular
fungi, then depleted atmospheric oxygen via decay processes and inhibited further
carbon accumulation (20, 31, 35). Biological underpinnings to a late Paleozoic
oxygen pulse are thus well-grounded, although more precise estimates of excur-
sions in atmospheric carbon dioxide and oxygen await further investigation.
Constraints on fire processes can be used to evaluate the validity of geophysical
reconstructions of ancient atmospheres. The models of atmospheric oxygen con-
sidered here suggest concentrations approaching but not exceeding 35%, a value
that represents an approximate threshold for sustained combustion of the terres-
trial biosphere (36, 37). Proliferation of terrestrial ecosystems would clearly be
precluded under such circumstances, and analysis of charcoal deposits and of the
vegetational fossil record demonstrates that atmospheric oxygen concentrations
have not exceeded this upper bound (38). At the lower extreme, the minimum
oxygen concentration required for contemporary natural fuels to ignite is about
13% (36, 39); the regular presence of charcoal in the fossil record suggests that
atmospheric oxygen never fell below this level (40?42). Methodologically, it is
important to realize that ignition of homogeneous woody or paper materials in
modern laboratory contexts is not functionally equivalent to the propagation of
fire in natural ecosystems. Extrapolation of experimental results with contempo-
rary fuels to the prediction of fire regimes in Paleozoic vegetation is necessarily
imprecise, although some inference is possible (34). In sum, these considerations
suggest that atmospheric oxygen levels have been historically constrained
between 13 and 35%; the best estimate derived from atmospheric modeling (15?
35%; Reference 21) is nominally congruent with this range.
No direct indicator is available to test empirically these predictions for ancient
levels of atmospheric oxygen. However, diverse geochemical and isotopic evi-
dence is consistent with substantial late Permian declines in both marine and
atmospheric oxygen concentrations, declines that likely contributed to the end-
Permian extinction event (43?51). Furthermore, atmospheric models incorporat-
ing effects of plant terrestrialization suggest substantial changes in the
concentration of carbon dioxide. Both modeling and empirical observations dem-
onstrate an approximate 10-fold reduction in carbon dioxide from the mid- to late
Paleozoic (22, 27?29, 53?56). For plants, the physiological effects of this draw-
down in carbon dioxide are evident in paleontological analyses of stomatal density
and other proxies of photosynthetic activity (56?58). Shortcomings of atmo-
spheric modeling notwithstanding, a diversity of empirical evidence is consistent
both with carbon dioxide drawdown and with an oxygen pulse in the mid- to late
Paleozoic, and with enhanced levels of oxygen in the mid- to late Mesozoic and
in much of the Cenozoic.
OXYGEN AND ANIMAL FLIGHT EVOLUTION 139
THE ORIGINS OF ANIMAL FLIGHT
Powered flapping flight evolved independently four times during the Phanerozoic,
once in the insects and three times among vertebrates. Evolutionary origins of all
volant taxa remain controversial, but both insects and the three taxa of flying
vertebrates may have originated during periods of atmospheric hyperoxia. Flight
is both biomechanically and physiologically demanding, and the evolution of
flight requires dramatic up-regulation of metabolic capacity in addition to the
expression of flapping structures that generate useful aerodynamic forces.
Although wings are not homologous between insects and volant vertebrates, com-
mon physical mechanisms of flight suggest that historically elevated concentra-
tions of atmospheric oxygen have yielded similar aerodynamic consequences
during flight evolution.
Wings and Aerodynamic Force Production
Fundamental to force generation by flapping animal wings is the physical nature
of the medium to which the wings transfer momentum (59). Air density, the mass
of air per unit volume, is a principal determinant of force production by airfoils,
biological and otherwise; aerodynamic forces on structures tend to increase lin-
early with air density. Also, air responds to applied forces by resisting not defor-
mation but rather the rate of deformation, as determined by the viscosity or the
internal resistance to flow. For an object such as a wing moving in a fluid, the
ensuing reaction force acting parallel to the direction of movement is termed drag.
Presence of the wing removes momentum and thus kinetic energy from the mov-
ing fluid system, and acts to slow the wing relative to flow. This pressure drag,
also known as inertial drag, derives from disruption of the inertial characteristics
of the moving flow field and varies with fluid density, relative fluid velocity, and
the wing?s shape.
Viscous drag, by contrast, emerges from boundary interactions between the
wing and the surrounding fluid. Viscosity ensures resistance to flow between
adjacent layers of fluid, engendering shear forces within the boundary layer over
the wing and imposing a reaction force throughout the fluid that terminates at the
wing?s surface. Viscous drag thus varies not with density but with fluid viscosity,
the relative fluid velocity, and the total wetted surface area of the wing. Pressure
and viscous drag have in common a dependence on relative fluid velocity and on
object dimensions, but differ fundamentally in their dependence on fluid density
and viscosity, respectively. The ratio of inertial to viscous forces on an object
moving within a fluid will therefore vary with the ratio of density and viscosity,
and must additionally change with object dimensions and its relative velocity.
The simplest dimensionless formulation of these four parameters is the Reynolds
number (Re), which is proportional to the ratio of inertial to viscous forces that
act on objects moving within fluids. Different situations of fluid flow are physi-
cally equivalent if the corresponding Res are approximately the same. In contrast,
140 DUDLEY
variation in either the density or viscosity of a fluid alters the relative magnitude
of inertial and viscous forces.
Just as drag is the force parallel to flow around an object, lift is defined as the
component of force orthogonal to flow and thus perpendicular to drag in a two-
dimensional perspective. For wings at low angles of attack, air flows smoothly
over both dorsal and ventral wing surfaces. Because pressure drag of well-
designed airfoils is low, momentum extraction from the flow field is minimized.
The positive camber of the wing and the downward deflection of air near the
trailing wing edge yield slightly different translational velocities for the dorsal
and ventral airstreams?airflow is slightly faster above and slower beneath the
airfoil. Bernoulli?s Principle indicates that this difference in dorsal and ventral
airstream velocities results in a pressure gradient; a net force (i.e. lift) acts dorsally
upon the wing. In a cross-sectional perspective, air appears to move anteriorly
from the ventral to dorsal wing surface and to yield a net rotational movement
of air around the wing. This motion of air is equivalent to a rotating flow field or
vortex that circulates around the wing and that is centered about the wing itself.
For every bound wing vortex, a starting vortex of comparable magnitude but of
opposite sense is shed into the wake, ensuring conservation of angular momentum
in the fluid.
The physical origin of lift thus lies within the creation of net air circulation
about a translating wing. As a wing translates in space, the pressure gradient
underlying lift production yields airflow not only around spanwise wing sections
(the bound circulation), but also around the wing tip, creating the so-called tip
vortex. The tip vortex is linked to the starting and the bound wing vortices and
creates a closed vortex loop that exerts a momentum flux on the surrounding air.
The downward momentum induced by the presence of the vortex loop is accord-
ingly proportional to the mass of the air moved ventrally (i.e. to the air density)
and to the velocity at which the air moves (i.e. the circulation). Lift production
thus increases with increased air density (59). The Re at which a wing is operating
also exerts a strong influence on lift production. Under highly viscous circum-
stances, vortex generation becomes more difficult as intermolecular stickiness
progressively impedes rotational motions of airflow. Circulatory lift becomes
increasingly more difficult to maintain at low Re, whereas the effects of viscous
drag also become more pronounced. The lift:drag ratio must then decrease at
lower Re, as confirmed empirically for insect and vertebrate wings (60, 61).
Lifting characteristics of wings are thus strongly dependent on both air density
and the Re at which they operate. Concomitant with large-scale changes in oxygen
concentration during the Phanerozoic, numerous physical characteristics of air
must vary if, as suggested previously, nitrogen content of the atmosphere remains
constant (23). Parameters such as heat capacity of air, thermal conductivity, dif-
fusivity, and the speed of sound would all have substantially changed under hype-
roxic conditions. Of particular interest for the evolution of animal flight is the
possibility of density-mediated increases in aerodynamic performance during ini-
tial periods of wing evolution (23, 25). The predicted value of air density at the
OXYGEN AND ANIMAL FLIGHT EVOLUTION 141
peak of the late Paleozoic oxygen pulse (285 MYa) is about 1.56 kg/m3, an
increase of 29% relative to the present sea-level value of 1.21 kg/m3. The relative
changes in the viscosity of air are, by contrast, much smaller (23). Winglets or
proto-wings moving in hyperoxic atmospheres would thus experience a higher
air density and would be moving at a comparably increased Re. Both conditions
would have facilitated lift production in early flying forms, although the magni-
tude of these aerodynamic effects would have varied both with morphological
features and with the particular wing and body motions of the taxon concerned.
Because insect and vertebrate wings are so distinct both ontogenetically and ana-
tomically, origins of flight in the two groups are considered separately.
Origins of Flight in Insects
The evolution of insect flight, a major event in biotic history, can be correlation-
ally linked to atmospheric hyperoxia. The origins of winged (pterygote) insects
are indeterminate but probably lie in the Upper Devonian or early Lower Car-
boniferous. Wingless hexapods are known from 395?390 MYa (62?65), whereas
fossils of pterygote hexapods (i.e. winged insects) date from approximately 325
MYa (66, 67). By the Upper Carboniferous, pterygotes are impressively diversi-
fied into about fifteen orders (68?71). Although pterygote insects are likely mono-
phyletic (72?78), the morphological origins of wings remain obscure. Wings have
been proposed to derive either from fixed paranotal outgrowths of thoracic and
abdominal segments in terrestrial taxa (79?86), or from ancestrally mobile gills,
gill covers, leg structures, or styli in aquatic forms (87?95). Unfortunately, no
transitional forms are known between the wingless apterygotes and the winged
pterygote insects, and the biology of early protopterygote forms remains specu-
lative and contentious. Of particular interest to the origins of flight is ancestral
habitat association of protopterygotes?were these animals terrestrial or aquatic?
Most evidence, particularly that relating to the physiology and origins of the
insect tracheal system, indicates that winged insects are derived from terrestrial
apterygote ancestors (61, 96?99). Aquatic larvae, particularly those of the extant
and phylogenetically basal Odonata and Ephemeroptera, appear to be secondarily
derived (73, 99, 100). Independent of habitat association, however, both larvae
and adults of ancestral winged insects probably expressed lateral lobed structures
on the abdominal as well as the thoracic segments (68, 90, 101?103). If winglets
or wings derived initially from fixed paranotal lobes, flapping motions might have
emerged indirectly through action of dorsoventral leg muscles that insert on the
thorax, as characterizes so-called bifunctional muscles in many extant insects
(104?108). A general question relating to wing origins concerns the possible
evolution of novel wing-like structures, as opposed to modification of pre-existing
morphological features (85). Acquisition of wings from ancestrally mobile struc-
tures might seem more parsimonious than the derivation of flapping wings from
stationary paranotal lobes, although the neontological and paleontological data
142 DUDLEY
available at present are insufficient to decide unequivocally between these two
hypotheses (61).
A variety of possible functional roles have been attributed to winglets or wings
of protopterygotes, including aerodynamic utility, epigamic display during court-
ship, and thermoregulation (109?115). Hydrodynamic use for what ultimately
became aerodynamic structures has been proposed for ancestrally aquatic proto-
pterygotes, as have been possibly amphibious lifestyles (116, 117). Protoptery-
gotes could also have used wing-like structures in air either to drift passively or
to skim actively along water surfaces, as do many extant insect taxa (118?122).
This behavior is probably a derived rather than a retained ancestral trait of winged
insects (61, 123, 124). Although improbable for reasons outlined above, biome-
chanical considerations suggest that aquatic protopterygotes would have been
unlikely to evolve wings that served aerodynamic functions. Water and air differ
by almost three orders of magnitude in density, with a corresponding difference
in the Re and in the nature of forces generated by oscillating structures. The
functionality of wing designs intermediate to either hydrodynamic or aerody-
namic force generation is correspondingly unclear (61). Forces of surface tension
would present a formidable physical barrier to partial body emergence as well as
to projection and oscillation of flattened structures, particularly for the body sizes
(2?4 cm) likely characteristic of ancestral pterygotes (64, 125, 126).
Given the assumption of terrestrial pterygote ancestors, a standard explanation
for the evolution of wings has been that these structures aerodynamically facilitate
jumping escapes from predators on land. Suggestively, a suite of morphological
and behavioral protoadaptations for jump-mediated glides are evident among
extant apterygote hexapods, the terrestrial sister taxon of the winged insects. Tho-
racic paranotal lobes as well as styli on the legs and abdominal segments of extant
apterygotes could potentially have served in ancestral taxa to generate lift and to
facilitate saltatorial escape (61). Neurobiological studies also support the ancestral
presence of dedicated sensorimotor pathways underlying escape behavior in both
apterygotes and pterygotes (127?129). The startle response of ancestral apterygote
insects was then apparently co-opted during pterygote evolution to stimulate
jumping, wing flapping, and even evasive flight once airborne (130?137). The
historical context of early pterygote evolution was appropriate for imposition of
intense predatory pressure by both invertebrates and vertebrates, with a diversity
of insectivorous arthropods (particularly arachnids), amphibians, and reptiles
found in Devonian and Carboniferous terrestrial ecosystems (138?142). Various
morphological characteristics evident among the early Upper Carboniferous
entomofauna are also consistent with the hypothesis of predatory defense (141,
143, 144).
Furthermore, the increasing arborescence and geometrical complexity of ter-
restrial vegetation through the Devonian and into the Carboniferous would have
provided a three-dimensional substrate suitable for jumping escapes and ulti-
mately the evolution of flight in protopterygotes (145). Terrestrial vegetation dur-
ing the period of pterygote origins was composed of fern-like lycopsids and
OXYGEN AND ANIMAL FLIGHT EVOLUTION 143
psilopsids, together with arborescent lycopods, sphenopsids, and progymno-
sperms (146?150). Lateral jumps from branches or leaves, and particularly jumps
directed downward, would have yielded the increased translational velocities nec-
essary for substantial buildup of lift on both thoracic and abdominal projections.
Access to nutritional resources, appropriate microhabitats, and suitable oviposi-
tion sites would have been facilitated by such an increased capacity for three-
dimensional movements. Increased feeding on plants would have been one
outcome of enhanced mobility. For example, non-insect hexapods (e.g. Collem-
bola) that feed on living plant tissue are rare, whereas approximately 85% of
extant insect species are phytophagous at some stage in their life cycle (151).
Most tellingly, the fossil record confirms feeding on plants by insects in the Upper
Carboniferous and early Permian (152?154).
Given this paleobiological background to wing evolution in insects, it is
instructive to consider two distinct features of atmospheric hyperoxia that may
have contributed to the evolution of flight. Biomechanically, the greater density
of a hyperoxic atmosphere would yield enhanced aerodynamic characteristics of
both winglets and the bodies of protopterygotes. Alternatively phrased, less wing-
let area would be necessary to generate comparable forces if air density were
higher, although the magnitude of this effect is strongly dependent on the nature
of vortex production around the body and winglets and on the relative air velocity
and/or acceleration during takeoffs. Aerodynamic forces on both the body and
winglets of protopterygotes would enhance aerial escape during either jumping
or steady-state glides (115, 125, 155). The de facto increase in Re associated with
increased air density may furthermore have contributed to greater lift production
by winglets, particularly for the low aspect ratio wings and Re relevant to ptery-
gote evolution (61, 156). In contrast to theories of wing evolution that require
aquatic protopterygotes and a discontinuous selection regime across the air-water
interface, escape jumping in terrestrial protopterygotes would from the outset
favor gradual and continuous improvement in lift generation. Winglet mobility
would then enhance force production, and additionally could be used to control
body orientation during flight and while landing (109, 125, 157, 158). Increased
air density and higher Re would have been advantageous in any evolutionary
scenario involving lift production by winglets or wings. Greater air densities may
also have facilitated flight in progressively larger taxa, leading to the giant pter-
ygotes of the late Paleozoic, as discussed below (61, 159).
Physiologically, an increased oxygen content of the late Paleozoic atmosphere
would have favored higher and ever more sustained levels of the oxidative metab-
olism required for flight. As wing flapping became more rapid and of greater
amplitude, the energetic costs of flight would have increased substantially; tho-
racic muscles of extant insects in flight exhibit the highest mass-specific rates of
oxygen consumption known for any locomotor tissue (61, 160, 161). This demand
for oxygen is met by the air-filled tracheal system, a branching network of cutic-
ular tubes that functions primarily through gaseous diffusion rather than by con-
vective air movement (162?165). Both diffusion and tracheal convection
144 DUDLEY
represent potentially rate-limiting steps in insect respiration, although the much
reduced thoracic diameters of small insects (e.g. Drosophila) probably alleviate
any metabolically based need for active ventilation of the tracheal system. In
larger insects, however, diffusional limits on oxygen supply likely constrain max-
imum body size. For example, flight metabolic rates in an extant dragonfly species
vary directly with ambient oxygen concentration (166), a result consistent with
diffusion-limited oxygen delivery by the tracheal system. In the late Paleozoic,
diffusion of oxygen to the respiring flight muscles would obviously have been
enhanced by greater partial pressures of this gas, as well as by the increased
diffusion constants associated with an increased total pressure of the atmosphere
(25).
Moreover, a spectacular bioindicator of atmospheric hyperoxia is provided by
gigantism among late Paleozoic terrestrial arthropods (23, 25, 26). Gigantism in
the Carboniferous and Permian was characteristic of many arthropod taxa, includ-
ing a number of phylogenetically unrelated pterygote orders (141, 167, 168) as
well as apterygote insects (169) and other arthropods, including diplopods, arthro-
pleurids, and arachnids (138, 141, 170?172). Although a number of non-mutually
exclusive selective forces may have contributed to the evolution of gigantism (90,
141, 173), relaxation of diffusional constraints on maximum body size through
atmospheric hyperoxia seems to be the most parsimonious explanation for this
phenomenon (19, 23, 25, 174, 175). Necessarily global effects of hyperoxia on
animal physiology were thus manifested in the form of dramatically increased
body size in diverse arthropod lineages. Furthermore, all giant arthropod taxa of
the late Paleozoic went extinct by mid- to end-Permian (23, 25, 168), as would
be predicted by the concomitant decline in atmospheric oxygen concentration.
Also consistent with the hypothesis linking atmospheric oxygen and diffusional
constraints on insect body size is a secondary peak of gigantism (e.g. among
mayflies) in the Cretaceous at a time of elevated oxygen concentrations (25, 168).
Systematic analysis of Paleozoic gigantism has been precluded to date because
of generally poor fossil preservation and incomplete knowledge of interspecific
body size distributions. However, detailed analysis of phyletic size change within
diffusion-limited taxa does represent an important bioassay for possible effects
of atmospheric hyperoxia in the late Paleozoic and the Cretaceous/Tertiary.
Origins of Vertebrate Flight
As with winged insects, the historical origins of flying vertebrates are known only
imprecisely. Nonetheless, circumstantial evidence is at least consistent with ori-
gins of bats, birds, and pterosaurs during times of hyperdense and hyperoxic
atmospheres. The most recent of volant vertebrates are the bats, with a modern
morphology apparent in a microchiropteran fossil from 50 MYa. Given this date,
chiropteran origins would appear to lie within the early Paleocene or late Creta-
ceous (176, 177). A similar appearance during hyperoxic times seems likely for
birds in the mid- to late Jurassic (178?180). The timing of pterosaur origination
OXYGEN AND ANIMAL FLIGHT EVOLUTION 145
and early diversification is unknown but possibly lies within the Permian given
well-developed pterosaur morphologies by the mid- to late Triassic (181, 182).
Because a detailed fossil record is unavailable for any of these three taxa, bio-
mechanical analysis of the transitional forms of flight is seriously constrained.
Independent of the behavioral or ecological context of flight, however, greater
atmospheric density would facilitate aerodynamic force production, whereas
increased oxygen partial pressures would similarly enhance oxygen transport to
and within the muscles powering flight. Aforementioned discussions of vertebrate
flight origins generally assume atmospheric composition of the present day,
although the implications of flight evolution in physically variable atmospheres
clearly deserve further attention. In a similar vein, reconstructions of ancestral
flight physiology that are based on comparisons with extant taxa (183?185) may
not fully reflect aerial performance of now extinct taxa in hyperoxic and hyper-
dense atmospheres.
Most scenarios of vertebrate flight evolution presume jumping takeoffs
(directed either with or against gravity), whereas jump-initiated gliding origins
of flight are biomechanically more parsimonious than ?ground-up? hypotheses
for both volant vertebrates and insects (60, 61, 186?188). Jumping via a startle
response is widespread among animals (189), and one potential commonality
among volant vertebrates and pterygote insects is acquisition of active flight via
the pathway of jumping and subsequent gliding to escape predation. Also,
unsteady aerodynamic performance during the accelerating portion of a jump may
be of greater relevance to escape success than steady-state glides (61). Jump-
initiated glides that increased survivorship during predation attempts would
potentially select for greater aerodynamic performance, an effect likely to be
enhanced at the higher Re of hyperdense air. That predation is a major factor
underlying the evolution of aerial locomotion is suggested by the increased lon-
gevity of flying animals relative to their non-volant counterparts. For example,
volant endotherms have mortality rates significantly lower than those for non-
volant endotherms of equivalent body mass (190?193). Paleophysiological sim-
ulations of jumping performance can be carried out in extant gliding and flying
taxa using variable-density gas mixtures (194), and potentially can be used to
investigate the ancestral role of jumping and takeoff aerodynamics in flight
evolution.
CONCLUSIONS AND IMPLICATIONS
Modeling, isotopic and geochemical data, and indirect biological indicators sug-
gest major historical changes in atmospheric composition, particularly during the
late Paleozoic but also in the late Mesozoic and Tertiary. Periods of atmospheric
hyperoxia can be correlationally linked with the late Paleozoic origin of flight in
insects and with contemporaneous arthropod gigantism and can be more tenta-
tively associated with the origin of flight in three vertebrate lineages. Causal
146 DUDLEY
association between periods of hyperoxia and the appearance of flight resides at
two levels: biomechanical enhancement of aerodynamic force production through
increases in air density and the Re, and physiological facilitation of the oxygen
flux required to sustain high rates of flight metabolism. In and of themselves,
increased oxygen concentrations did not act directly on particular phenotypes or
behaviors. Instead, a new selective background of slowly increasing oxygen con-
centrations provided the opportunity (but not necessity) for the evolution and
progressive enhancement of a novel locomotor capacity (26).
Falsification of such historical hypotheses is necessarily challenging, but neon-
tological investigations of animal flight performance provide a useful perspective
on paleobiological reconstructions. Postulated air densities and oxygen concen-
trations of ancient atmospheres can be mimicked experimentally using different
combinations of oxygen and nonreactive gases (194). Similar gas mixtures can
be used to investigate the effects of atmospheric hyperoxia on the biomechanics
and physiology of flight in extant volant forms. Phenotypic plasticity in the res-
piratory system of insects (195?197) and vertebrates (198?204) suggests wide-
spread occurrence of ontogenetic as well as short-term and chronic physiological
flexibility in response to variable oxygen availability. Over evolutionary time
scales, artificial selection for enhanced flight performance in Drosophila (205)
can be imposed in different gas mixtures that decouple the effects of hyperoxia
from those associated with a hyperdense flight medium (25, 194).
In modern ecological contexts, variation in altitude provides the closest ana-
logue to historic variation in atmospheric composition. Covariance in air density
and oxygen partial pressure characterizes altitudinal gradients, as would similarly
characterize atmospheres with varying oxygen content and a constant quantity of
nitrogen. Modern volant taxa display remarkable flight abilities under both hyp-
oxic and hypodense conditions (61, 206?209). In natural montane contexts, wing
length of birds and insects varies directly with altitude, suggesting morphological
compensation for the greater power expenditure during flight at lower air densities
(194, 210?213). The wide range of extant adaptations to variation in air density
and oxygen availability suggests substantial evolutionary capacity to respond to
variable atmospheric composition. Although considerable uncertainty still sur-
rounds paleoatmospheric reconstructions, further study of animal flight perfor-
mance in simulated ancient atmospheres is likely to prove rewarding.
ACKNOWLEDGMENTS
I thank Doug Altshuler, Bob Berner, Carl Gans, and Tim Lenton for useful com-
ments, and acknowledge grants from the National Science Foundation for
research support.
Visit the Annual Reviews home page at www.AnnualReviews.org.
OXYGEN AND ANIMAL FLIGHT EVOLUTION 147
LITERATURE CITED
1. Hudson JD. 1989. Palaeoatmospheres in
the Phanerozoic. J. Geol. Soc. London
146:155?60
2. Lambert IB, Donnelly TH. 1991. Atmo-
spheric oxygen levels in the Precam-
brian: a review of isotopic and geological
evidence. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 97:83?91
3. Des Marais DJ, Strauss H, Summons RE,
Hayes, JM. 1992. Carbon isotope evi-
dence for the stepwise oxidation of the
Proterozoic environment. Nature 359:
605?9
4. Knoll AH. 1992. The early evolution of
eukaryotes: a geological perspective. Sci-
ence 256:622?27
5. Canfield DE, Teske A. 1996. Late Pro-
terozoic rise in atmospheric oxygen con-
centration inferred from phylogenetic
and sulphur-isotope studies. Nature
382:127?32
6. Gilbert DL. 1996. Evolutionary aspects
of atmospheric oxygen and organisms.
In Environmental Physiology, ed. MJ
Fregly, CM Blatteis, 2:1059?94. New
York: Oxford Univ. Press
7. Nursall JR. 1959. Oxygen as a prereq-
uisite to the origin of Metazoa. Nature
183:1170?72
8. Berkner LV, Marshall LC. 1965. On the
origin and rise of oxygen concentration
in the Earth?s atmosphere. J. Atmos. Sci.
22:225?61
9. Berkner LV, Marshall LC. 1967. The rise
of oxygen in the earth?s atmosphere with
notes on the martian atmosphere. Adv.
Geophys. 12:309?31
10. Knoll AH, Hayes JM, Kaufman AJ,
Swett K, Lambert, IB. 1986. Secular
variation in carbon isotope ratios from
Upper Proterozoic successions of Sval-
bard and East Greenland. Nature
321:832?38
11. Conway Morris S. 1993. The fossil rec-
ord and the early evolution of the Meta-
zoa. Nature 361:219?25
12. Kasting JF, Holland HD, Kump LR.
1992. Atmospheric evolution: the rise of
oxygen. In The Proterozoic Biosphere. A
Multidisciplinary Study, ed. JW Schopf,
C Klein, pp. 159?63. New York: Cam-
bridge Univ. Press
13. Kasting JF. 1993. Earth?s early atmo-
sphere. Science 259:920?26
14. Logan GA, Hayes JM, Hieshima GB,
Summons RE. 1995. Terminal Protero-
zoic reorganization of biogeochemical
cycles. Nature 376:53?56
15. Holland HD. 1984. The Chemical Evo-
lution of the Atmosphere and Oceans.
Princeton, NJ: Princeton Univ. Press
16. Budyko MI, Ronov AB, Yanshin AL.
1985. History of the Earth?s Atmosphere.
Berlin: Springer-Verlag
17. Schlesinger WH. 1991. Biogeochemis-
try: An Analysis of Global Change. San
Diego: Academic
18. Deleted in proof
19. Tappan H. 1974. Molecular oxygen and
evolution. In Molecular Oxygen in Biol-
ogy, ed. O. Hayaishi, pp. 81?135. Am-
sterdam: North Holland
20. Robinson JM. 1991. Phanerozoic atmo-
spheric reconstructions: a terrestrial per-
spective. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 97:51?62
21. Berner RA, Canfield DE. 1989. A new
model for atmospheric oxygen over
Phanerozoic time. Am. J. Sci. 289:333?
61
22. Berner RA. 1998. The carbon cycle and
CO2 over Phanerozoic time: the role of
land plants. Philos. Trans. R. Soc. Lon-
don Ser. B 353:75?82
23. Graham JB, Dudley R, Aguilar N, Gans
C. 1995. Implications of the late Palaeo-
zoic oxygen pulse for physiology and
evolution. Nature 375:117?20
148 DUDLEY
24. Graham JB, Aguilar N, Dudley R, Gans,
C. 1997. The late Paleozoic atmosphere
and the ecological and evolutionary
physiology of tetrapods. In Amniote Ori-
gins: Completing the Transition to Land,
ed. SS Sumida, KLM Martin, pp. 141?
67. New York: Academic
25. Dudley R. 1998. Atmospheric oxygen,
giant Paleozoic insects and the evolution
of aerial locomotor performance. J. Exp.
Biol. 201:1043?50
26. Gans C, Dudley, R, Aguilar NM, Graham
JB. 1999. Late Paleozoic atmospheres
and biotic evolution. Hist. Biol. 13:199?
219
27. Berner RA. 1990. Atmospheric carbon
dioxide levels over Phanerozoic time.
Science 249:1382?85
28. Berner RA. 1994. GEOCARB II: a
revised model of atmospheric CO2 over
Phanerozoic time. Am. J. Sci. 294:56?91
29. Berner RA. 1997. The rise of plants and
their effect on weathering and atmo-
spheric CO2. Science 276:544?46
30. Mackenzie FT, Morse JW. 1992. Sedi-
mentary carbonates through Phanerozoic
time. Geochim. Cosmochim. Acta
56:3281?95
31. Visscher H, Brinkhuis H, Dilcher DL,
Elsik WC, Eshet Y, et al. 1996. The ter-
minal Paleozoic fungal event: evidence
of terrestrial ecosystem destabilization
and collapse. Proc. Natl. Acad. Sci. USA
93:2155?58
32. Algeo TJ, Scheckler SE. 1998. Terres-
trial-marine teleconnections in the Devo-
nian: links between the evolution of land
plants, weathering processes, and marine
anoxic events. Philos. Trans. R. Soc.
London Ser. B 353:113?30
33. Van Cappellen P, Ingall ED. 1996. Redox
stabilization of the atmosphere and
oceans by phosphorus-limited marine
productivity. Science 271:493?96
34. Lenton TM, Watson AJ. 1999. Redfield
revisited: 2. What regulates the oxygen
content of the atmosphere? Global Bio-
geochem. Cycles. In press
35. Robinson JM. 1990. Lignin, land plants,
and fungi: biological evolution affecting
Phanerozoic oxygen balance. Geology
15:607?10
36. Watson A, Lovelock JE, Margulis L.
1978. Methanogenesis, fires and the
regulation of atmospheric oxygen. Bio-
Systems 10:293?98
37. Kump LR. 1989. Chemical stability of
the atmosphere and ocean. Global
Planet. Change 1:12 3?26
38. Jones TP, Chaloner WG. 1991. Fossil
charcoal, its recognition and palaeoat-
mospheric significance. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 97:39?50
39. Rasbash DJ, Langford B. 1968. Burning
of wood in atmospheres of reduced oxy-
gen concentrations. Combust. Flame 12:
33?40
40. Cope MJ, Chaloner WG. 1980. Fossil
charcoal as evidence of past atmospheric
composition. Nature 283:647?49
41. Clark FRS, Russell DA. 1981. Fossil
charcoal and the palaeoatmosphere.
Nature 290:428
42. Chaloner WG. 1989. Fossil charcoal as
an indicator of palaeoatmospheric oxy-
gen level. J. Geol. Soc. London 146:
171?74
43. Berner RA. 1989. Drying, O2 and mass
extinction. Nature 340:603?4
44. Holser WT, Scho?nlaub HP, Attrep M Jr,
Boeckelmann K, Klein P, et al. 1989. A
unique geochemical record at the Per-
mian/Triassic boundary. Nature 337:
39?44
45. Ma?kowski K, Gruszczyn?ski M, Hoff-
man A, Halas S. 1989. Oceanic stable
isotope composition and a scenario for
the Permo-Triassic crisis. Hist. Biol.
2:289?309
46. Hallam A. 1991. Why was there a
delayed radiation after the end-Palaeo-
zoic extinctions? Hist. Biol. 5:257?62
47. Wignall PB, Hallam A. 1992. Anoxia as
a cause of the Permian/Triassic mass
extinction: facies evidence from northern
Italy and the western United States.
OXYGEN AND ANIMAL FLIGHT EVOLUTION 149
Palaeogeogr. Palaeoclimatol. Palaeoe-
col. 93:21?46
48. Erwin DH. 1993. The Great Paleozoic
Crisis: Life and Death in the Permian.
New York: Columbia Univ. Press
49. Wignall PB, Hallam A. 1993. Griesba-
chian (Earliest Triassic) palaeoenviron-
mental changes in the Salt Range,
Pakistan and southeast China and their
bearing on the Permo-Triassic mass
extinction. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 102:215?37
50. Wignall PB, Twitchett RJ. 1996. Oceanic
anoxia and the End Permian mass extinc-
tion. Science 272:1155?58
51. Isozaki Y. 1997. Permo-Triassic bound-
ary superanoxia and stratified supero-
cean: records from lost deep sea. Science
276:235?38
52. Deleted in proof
53. Berner RA. 1991. A model for atmo-
spheric CO2 over Phanerozoic time. Am.
J. Sci. 291:339?76
54. Mora CI, Driese SG, Colarusso LA.
1996. Middle to late Paleozoic atmo-
spheric CO2 levels from soil carbonate
and organic matter. Science 271:1105?7
55. Retallack GJ. 1997. Early forest soils and
their role in Devonian global change. Sci-
ence 276:583?85
56. McElwain JC. 1998. Do fossil plants sig-
nal palaeoatmospheric CO2 concentra-
tion in the geological past? Philos. Trans.
R. Soc. London Ser. B 353:83?96
57. Beerling DJ, Woodward FI, Lomas MR,
Wills MA, Quick WP, Valdes PJ. 1998.
The influence of Carboniferous palaeo-
atmospheres on plant function: an expe-
rimental and modelling assessment.
Philos. Trans. R. Soc. London Ser. B
353:131?40
58. Edwards D. 1998. Climate signals in
Palaeozoic land plants. Philos. Trans. R.
Soc. London Ser. B 353:141?57
59. Vogel S. 1994. Life in Moving Fluids:
The Physical Biology of Flow. Princeton,
NJ: Princeton Univ. Press
60. Norberg UM. 1990. Vertebrate Flight.
Berlin: Springer-Verlag
61. Dudley R. 2000. The Biomechanics of
Insect Flight: Form, Function, Evolution.
Princeton, NJ: Princeton Univ. Press
62. Whalley P, Jarzembowski EA. 1981. A
new assessment of Rhyniella, the earliest
known insect, from the Devonian of
Rhynie, Scotland. Nature 291:317
63. Shear WA, Grierson JD, Rolfe WDI,
Smith EL, Norton RA. 1984. Early land
animals in North America: evidence
from Devonian age arthropods from Gil-
boa, New York. Science 224:492?94
64. Labandeira CC, Beall BS, Hueber FM.
1988. Early insect diversification: evi-
dence from a Lower Devonian bristletail
from Que?bec. Science 242:913?16
65. Jeram AJ, Selden PA, Edwards D. 1990.
Land animals in the Silurian: arachnids
and myriapods from Shropshire,
England. Science 250:658?61
66. Nelson CR, Tidwell WD. 1987. Brodiop-
tera stricklani n.sp. (Megasecoptera:
Brodiopteridae), a new fossil insect from
the Upper Manning Canyon Shale For-
mation, Utah (lowermost Namurian B).
Psyche 94:309?16
67. Brauckmann C, Zessin W. 1989. Neue
Meganeuridae aus dem Namurian von
Hagen-Vorhalle (BRD) und die Phylo-
genie der Meganisoptera. Dtsch. Ento-
mol. Z. N.F 36:177?215
68. Wootton RJ. 1981. Palaeozoic insects.
Annu. Rev. Entomol. 26:319?44
69. Wootton RJ. 1990. Major insect radia-
tions. In Major Evolutionary Radiations,
ed. PD Taylor, GP Larwood, Systematics
Assoc. Special Volume 42:187?208.
Oxford, UK: Clarendon
70. Kukalova?-Peck J. 1991. Fossil history
and the evolution of hexapod structures.
In The Insects of Australia, ed. CSIRO,
1:141?79. Ithaca, NY: Cornell Univ.
Press. 2nd ed.
71. Labandeira CC, Sepkoski JJ. 1993.
Insect diversity in the fossil record. Sci-
ence 261:310?15
150 DUDLEY
72. Boudreaux BH. 1979. Arthropod Phylog-
eny with Special Reference to Insects.
New York: Wiley & Sons
73. Hennig W. 1981. Insect Phylogeny.
Chichester, UK: Wiley & Sons
74. Kristensen NP. 1981. Phylogeny of
insect orders. Annu. Rev. Entomol.
26:135?57
75. Kristensen NP. 1989. Insect phylogeny
based on morphological evidence. In The
Hierarchy of Life, ed. B Fernholm, K
Bremer, H. Jo?rnvall, pp. 295?306.
Amsterdam: Elsevier Sci.
76. Kristensen NP. 1991. Phylogeny of
extant hexapods. See Ref. 70, pp. 125?
40
77. Kristensen NP. 1997. The groundplan
and basal diversification of the hexapods.
In Arthropod Relationships, ed. RA For-
tey, RH Thomas, pp. 281?93. London:
Chapman & Hall
78. Brodsky AK. 1994. The Evolution of
Insect Flight. Oxford, UK: Oxford Univ.
Press
79. Crampton GC. 1916. The phylogenetic
origin and the nature of the wings of
insects according to the paranotal theory.
J. NY Entomol. Soc. 24:1?39
80. Forbes WTM. 1943. The origin of wings
and venational types in insects. Am.
Midl. Nat. 29:381?405
81. Hamilton KGA. 1971. The insect wing,
Part I. Origin and development of wings
from notal lobes. J. Kans. Entomol. Soc.
44:421?33
82. Rasnitsyn AP. 1981. A modified paran-
otal theory of insect wing origin. J. Mor-
phol. 168:331?38
83. Quartau JA. 1985. On some objections to
the paranotal theory on the origin of
insect wings. Bol. Soc. Port. Entomol.
Suppl. 1:359?71
84. Quartau JA. 1986. An overview of the
paranotal theory on the origin of the insect
wings. Publ. Inst. Zool. Dr. Augusto
Nobre, Fac. Cienc. Porto 194:1?42
85. Wootton RJ. 1986. The origin of insect
flight: Where are we now? Antenna
10:82?86
86. Bitsch J. 1994. The morphological
groundplan of Hexapoda: critical review
of recent concepts. Ann. Soc. Entomol.
Fr. 30:103?29
87. Woodworth CW. 1906. The wing veins
of insects. Univ. Calif. Bull. Entomol.
1:1?52
88. Wigglesworth VB. 1973. Evolution of
insect wings and flight. Nature 246:127?
29
89. Wigglesworth VB. 1976. The evolution
of insect flight. In Insect Flight, ed. RC
Rainey, pp. 255?69. Oxford, UK:
Blackwell
90. Kukalova?-Peck J. 1978. Origin and evo-
lution of insect wings and their relation
to metamorphosis, as documented by the
fossil record. J. Morphol. 156:53?126
91. Kukalova?-Peck J. 1983. Origin of the
insect wing and wing articulation from
the arthropodan leg. Can. J. Zool.
61:1618?69
92. Kukalova?-Peck J. 1992. The ?Uniramia?
do not exist: the ground plan of the
Pterygota as revealed by Permian Dia-
phanopterodea from Russia (Insecta:
Paleodictyopteroidea). Can. J. Zool. 70:
236?55
93. Kukalova?-Peck J. 1997. Arthropod phy-
logeny and ?basal? morphological struc-
tures. See Ref. 77, pp. 249?68
94. Kukalova?-Peck J. 1997. Mazon Creek
insect fossils: the origin of insect wings
and clues about the origin of insect met-
amorphosis. In Richardson?s Guide to
the Fossil Fauna of Mazon Creek, ed.
CW Shabic, AA Hay, pp. 194?207. Chi-
cago: Northeastern Illinois Univ. Press
95. Averof M, Cohen, SM. 1997. Evolution-
ary origin of insect wings from ancestral
gills. Nature 385:627?30
96. Little C. 1983. The Colonisation of Land.
Origins and Adaptations of Terrestrial
Animals. Cambridge, UK: Cambridge
Univ. Press
97. Resh VH, Solem JO. 1984. Phylogenetic
OXYGEN AND ANIMAL FLIGHT EVOLUTION 151
relationships and evolutionary adapta-
tions of aquatic insects. In An Introduc-
tion to the Aquatic Insects of North
America, ed. RW Merrit, KW Cummins,
pp. 66?75. Dubuque, IA: Kendall/Hunt.
2nd ed.
98. Messner B. 1988. Sind die Insekten pri-
ma?re oder sekunda?re Wasserbewohner?
Dtsch. Entomol. Z. N.F 35:355?60
99. Pritchard G, McKee MH, Pike EM,
Scrimgeour GJ, Zloty J. 1993. Did the
first insects live in water or in air? Biol.
J. Linn. Soc. 49:31?44
100. Hinton HE. 1968. Spiracular gills. Adv.
Insect Physiol. 5:65?161
101. Carpenter FM. 1966. The Lower Permian
insects of Kansas. Part 11. The orders
Protorthoptera and Orthoptera. Psyche
73:46?88
102. Kukalova? J. 1968. Permian mayfly
nymphs. Psyche 75:310?27
103. Carroll SB, Weatherbee SD, Langeland
JA. 1995. Homeotic genes and the regu-
lation and evolution of insect wing num-
ber. Nature 375:58?61
104. Becker EG. 1952. The problem of the
origin and development of the wing in
insects. Chapter 1. Precursors of the
insect wing. Vestn. Mosk. Univ. Ser. Phys.
Math. Nat. Sci. 9:59?68
105. Wilson DM. 1962. Bifunctional muscles
in the thorax of grasshoppers. J. Exp.
Biol. 39:669?77
106. Ewer DW. 1963. On insect flight. J.
Entomol. Soc. S. Afr. 26:3?13
107. Ewer DW, Nayler LS. 1967. The ptero-
thoracic musculature of Deropeltis ery-
throcephala, a cockroach with a wingless
female, and the origin of wing move-
ments in insects. J. Entomol. Soc. S. Afr.
30:18?33
108. Fourtner CR, Randall JB. 1982. Studies
on cockroach flight: the role of continu-
ous neural activation of non-flight mus-
cles. J. Exp. Zool. 221:143?54
109. Wigglesworth VB. 1963. The origin of
flight in insects. Proc. R. Entomol. Soc.
London 28:23?32
110. Alexander RD, Brown WL. 1963. Mat-
ing behavior and the origin of insect
wings. Occas. Pap. Mus. Zool. Univ.
Mich. 628:1?19
111. Alexander RD. 1964. The evolution of
mating behaviour in arthropods. Symp. R.
Entomol. Soc. London 2:78?94
112. Whalley PS. 1979. New species of Pro-
torthoptera and Protodonata (Insecta)
from the Upper Carboniferous of Britain,
with a comment on the origin of wings.
Bull. Br. Mus. Nat. Hist. 32:85?90
113. Douglas MM. 1981. Thermoregulatory
significance of thoracic lobes in the evo-
lution of insect wings. Science 211:84?
86
114. Kingsolver JG, Koehl MAR. 1985. Aer-
odynamics, thermoregulation, and the
evolution of insect wings: differential
scaling and evolutionary change. Evolu-
tion 39:488?504
115. Kingsolver JG, Koehl MAR. 1994.
Selective factors in the evolution of
insect wings. Annu. Rev. Entomol. 39:
425?51
116. Bradley, JC. 1942. The origin and sig-
nificance of metamorphosis and wings
among insects. In Proc. 8th Am. Sci.
Congr. pp. 303?9. Washington DC: Dept.
State
117. Mamayev BM. 1977. The gravitational
hypothesis of the origin of insects. Ento-
mol. Rev. 54:13?17
118. Marden JH, Kramer MG. 1994. Surface-
skimming stoneflies: a possible interme-
diate stage in insect flight evolution.
Science 266:427?30
119. Marden JH, Kramer MG. 1995. Loco-
motor performance of insects with rudi-
mentary wings. Nature 377:332?34
120. Thomas ALR, Norberg RA? . 1996. Skim-
ming the surface?the origin of flight in
insects? Trends Ecol. Evol. 11:187?88
121. Samways MJ. 1994. ?Sailing? on the
water surface by adult male Enallagma
nigridorsum Selys (Zygoptera: Coena-
grionidae). Odonatologica 23:175?78
152 DUDLEY
122. Kramer MG, Marden JH. 1997. Almost
airborne. Nature 385:403?4
123. Will KW. 1995. Plecopteran surface-
skimming and insect flight evolution.
Science 270:1684?85
124. Ruffieux L, Elouard J-M, Sartori M.
1998. Flightlessness in mayflies and its
relevance to hypotheses on the origin of
insect flight. Proc. R. Soc. London Ser. B
265:2135?40
125. Flower JW. 1964. On the origin of flight
in insects. J. Insect Physiol. 10:81?88
126. Wootton RJ. 1976. The fossil record and
insect flight. See Ref. 89, pp. 235?54
127. Ritzmann RE, Fourtner CR. 1980. Flight
activity initiated via giant interneurons of
the cockroach: evidence for bifunctional
trigger interneurons. Science 210:443?45
128. Ritzmann RE. 1984. The cockroach
escape response. In Neural Mechanisms
of Startle Behavior, ed. RC Eaton, pp.
93?131. New York: Plenum
129. Edwards JS, Reddy GR. 1986. Mechan-
osensory appendages in the firebrat
(Thermobia domestica, Thysanura): a
prototype system for terrestrial predator
invasion. J. Comp. Neurol. 243:535?46
130. Bristowe WS. 1958. The World of Spi-
ders. London: Collins
131. Edwards JS. 1985. Predator evasion and
the origin of insect flight: an exercise in
evolutionary neuroethology. Soc. Neu-
rosci. Abstr. 11:497
132. Edwards JS, Palka J. 1991. Insect neural
evolution?a fugue or an opera? Semin.
Neurosci. 3:391?98
133. Edwards JS. 1992. Giant interneurons
and the origin of insect flight. In Nervous
Systems: Principles of Design and Func-
tion, ed. RN Singh, pp. 485?95. Bombay:
Wiley Eastern
134. Edwards JS. 1997. The evolution of
insect flight: implications for the evolu-
tion of the nervous system. Brain Behav.
Evol. 50:8?12
135. Ganihar D, Libersat F, Wendler G,
Camhi, JM. 1994. Wind-evoked evasive
responses in flying cockroaches. J.
Comp. Physiol. A 175:49?65
136. Kutsch W, Breidbach O. 1994. Homolo-
gous structures in the nervous systems of
Arthropoda. Adv. Insect Physiol. 24:1?11
137. Libersat F. 1994. The dorsal giant inter-
neurons mediate evasive behavior in fly-
ing cockroaches. J. Exp. Biol. 197:405?
11
138. Rolfe WD. 1980. Early invertebrate ter-
restrial fossils. In The Terrestrial Envi-
ronment and the Origin of Land
Vertebrates, ed. AL Panchen, Systemat-
ics Assoc. Special Vol. No. 15, pp. 117?
57. London: Academic
139. Rolfe WD. 1985. Early terrestrial arthro-
pods: a fragmentary record. Philos.
Trans. R. Soc. London Ser. B 309:207?
18
140. Carroll RL. 1988. Vertebrate Paleontol-
ogy and Evolution. New York: Freeman
141. Shear WA, Kukalova?-Peck J. 1990. The
ecology of Paleozoic terrestrial arthro-
pods: the fossil evidence. Can. J. Zool.
68:1807?34
142. Behrensmeyer AK, Damuth JD, Di-
Michele WA, Sues H-D, Wing SL. 1992.
Terrestrial Ecosystems through Time:
Evolutionary Paleoecology of Terrestrial
Plants and Animals. Chicago: Univ. Chi-
cago Press
143. Carpenter FM. 1971. Adaptations among
Paleozoic insects. Proc. N. Am. Paleon-
tol. Conv. 1969:1236?51
144. Burnham L. 1983. Studies on Upper Car-
boniferous insects: 1. The Geraridae
(Order Protorthoptera). Psyche 90:1?57
145. Kevan PG, Chaloner WG, Savile DBO.
1975. Interrelationships of early terres-
trial arthropods and plants. Paleontology
18:391?417
146. Chaloner WG, Sheerin A. 1979. Devo-
nian macrofloras. Spec. Pap. Palaeontol.
23:145?61
147. Gensel PG, Andrews HN. 1984. Plant
Life in the Devonian. New York: Praeger
148. Stewart WN. 1983. Paleobotany and the
Evolution of Plants. Cambridge, UK:
Cambridge Univ. Press
149. Raven JA. 1986. Evolution of plant life
forms. In On the Economy of Plant Form
OXYGEN AND ANIMAL FLIGHT EVOLUTION 153
and Function, ed. TJ Givnish, pp. 421?
92. Cambridge, UK: Cambridge Univ.
Press
150. Kenrick P, Crane PR. 1997. The origin
and early evolution of plants on land.
Nature 389:33?39
151. Strong DR, Lawton JH, Southwood R.
1984. Insects on Plants. Oxford, UK:
Blackwell
152. Rasnitsyn AP, Krassilov VA. 1996. First
find of pollen grains in the gut of Permian
insects. Paleontol. J. 30:484?90
153. Labandeira CC, Phillips TL. 1996. A
Carboniferous insect gall: insight into
early ecologic history of the Holometa-
bola. Proc. Natl. Acad. Sci. USA
93:8470?74
154. Labandeira CC, Phillips TL. 1996. Insect
fluid-feeding on Upper Pennsylvanian
tree ferns (Palaeodictyoptera, Marat-
tiales) and the early history of the
piercing-and-sucking functional feeding
group. Ann. Entomol. Soc. Am. 89:157?
83
155. Ellington CP. 1991. Aerodynamics and
the origin of insect flight. Adv. Insect
Physiol. 23:171?210
156. Ennos AR. 1989. The effect of size on
the optimal shapes of gliding insects and
seeds. J. Zool. London 219:61?69
157. Haupt H. 1941. Die a?ltesten geflu?gelten
Insekten und ihre Beziehungen zur Fauna
der Jetztzeit. Z. Naturwiss. (Halle)
94:60?121
158. Smart J. 1971. Palaeoecological factors
affecting the origin of winged insects.
Proc. 13th Int. Congr. Entomol. I:304?6
159. Harle? E? , Harle? A. 1911. Le vol de grands
reptiles et insectes disparus semble indi-
quer une pression atmosphe?rique e?leve?e.
Bull. Soc. Geol. Fr. Ser. 4 11:18?21
160. Kammer AE, Heinrich B. 1978. Insect
flight metabolism. Adv. Insect Physiol.
13:133?228
161. Casey TM. 1989. Oxygen consumption
during flight. In Insect Flight, ed. GJ
Goldsworthy, CH Wheeler, pp. 257?72.
Boca Raton, FL: CRC Press
162. Weis-Fogh T. 1964. Diffusion in insect
wing muscle, the most active tissue
known. J. Exp. Biol. 41:229?56
163. Weis-Fogh T. 1964. Functional design of
the tracheal system of flying insects as
compared with the avian lung. J. Exp.
Biol. 41:207?27
164. Kestler P. 1984. Respiration and respi-
ratory water loss. In Environmental Phys-
iology and Biochemistry of Insects, ed.
KH Hoffman, pp. 137?83. Berlin:
Springer-Verlag
165. Mill PJ. 1985. Structure and physiology
of the respiratory system. In Comprehen-
sive Insect Physiology, Biochemistry, and
Pharmacology, ed. GA Kerkut, LI Gil-
bert, 3:517?93. Oxford, UK: Pergamon
166. Harrison JF, Lighton JRB. 1998. Oxy-
gen-sensitive flight metabolism in the
dragonfly Erythemis simplicicollis. J.
Exp. Biol. 201:1739?44
167. Kukalova?-Peck J. 1985. Ephemeroid
wing venation based upon new gigantic
Carboniferous mayflies and basic mor-
phology, phylogeny, and metamorphosis
of pterygote insects (Insecta, Ephemer-
ida). Can. J. Zool. 63:933?55
168. Carpenter FM. 1992. Treatise on Inver-
tebrate Paleontology. Part R, Arthropoda
4, Volumes 3 and 4 (Hexapoda).
Lawrence, KS: Univ. Kansas Press
169. Kukalova?-Peck J. 1987. New Carbonif-
erous Diplura, Monura, and Thysanura,
the hexapod ground plan, and the role of
thoracic lobes in the origin of wings
(Insecta). Can. J. Zool. 65:2327?45
170. Rolfe WDI. 1969. Arthropleurida. In
Treatise on Invertebrate Paleontology.
Arthropoda 4: Crustacea (Except Ostra-
coda): Myriapoda, ed. RC Moore, C
Teichert, pp. 607?20. Lawrence, KS:
Univ. Kansas Press
171. Kraus O. 1974. On the morphology of
Paleozoic diplopods. Symp. Zool. Soc.
London 32:13?22
172. Briggs, DEG. 1985. Gigantism in
Palaeozoic arthropods. Spec. Pap. Pale-
ontol. 33:157
154 DUDLEY
173. Vermeij GJ. 1987. Evolution and Esca-
lation. Princeton, NJ: Princeton Univ.
Press
174. Rutten MG. 1966. Geologic data on
atmospheric history. Paleogeogr. Palaeo-
climatol. Palaeoecol. 2:47?57
175. Schidlowski M. 1971. Probleme der
atmospha?rischen Evolution im Pra?kam-
brium. Geol. Rundsch. 60:1351?84
176. Jepsen GL. 1970. Bat origins and evo-
lution. In Biology of Bats, ed. WA Wim-
satt, 1:1?64. London: Academic
177. Altringham JD. 1996. Bats: Biology and
Behaviour. Oxford, UK: Oxford Univ.
Press
178. Chiappe LM. 1995. The first 85 million
years of avian evolution. Nature
378:349?55
179. Qiang J, Currie PJ, Norell MA, Shu-An
J. 1998. Two feathered dinosaurs from
northeastern China. Nature 393:753?61
180. Padian K, Chiappe LM. 1998. The origin
and early evolution of birds. Biol. Rev.
73:1?42
181. Wild R. 1984. Flugsaurier aus der Ober-
trias von Italien. Naturwissenschaften
71:1?11
182. Wellnhofer P. 1991. The Illustrated
Encyclopedia of Pterosaurs. London:
Salamander
183. Ruben J. 1991. Reptilian physiology and
the flight capacity of Archaeopteryx.
Evolution 45:1?17
184. Ruben J. 1993. Powered flight in Archae-
opteryx: response to Speakman. Evolu-
tion 47:935?38
185. Speakman JR. 1993. Flight capabilities
in Archaeopteryx. Evolution 41:336?40
186. Norberg UM. 1985. Evolution of verte-
brate flight: an aerodynamic model for
the transition from gliding to active
flight. Am. Nat. 126:303?27
187. Rayner JMV. 1988. The evolution of ver-
tebrate flight. Biol. J. Linn. Soc. 34:269?
87
188. Bennett SC. 1997. The arboreal leaping
theory of the origin of pterosaur flight.
Hist. Biol. 12:265?90
189. Eaton RC, ed. 1984. Neural Mechanisms
of Startle Behavior. New York: Plenum
190. Tuttle MD, Stevenson D. 1982. Growth
and survival of bats. In Ecology of Bats,
ed. TH Kunz, pp. 105?50. New York:
Plenum
191. Pomeroy D. 1990. Why fly? The possible
benefits for lower mortality. Biol. J. Linn.
Soc. 40:53?65
192. Rose MR. 1991. Evolutionary Biology of
Aging. New York: Oxford Univ. Press
193. Holmes DJ, Austad SN. 1995. The evo-
lution of avian senescence patterns:
implications for understanding primary
aging processes. Am. Zool. 35:307?17
194. Dudley R, Chai P. 1996. Animal flight
mechanics in physically variable gas
mixtures. J. Exp. Biol. 199:1881?85
195. Loudon C. 1988. Development of Tene-
brio molitor in low oxygen levels. J.
Insect Physiol. 34:97?103
196. Loudon C. 1989. Tracheal hypertrophy
in mealworms: design and plasticity in
oxygen supply systems. J. Exp. Biol.
147:217?35
197. Greenberg S, Ar A. 1996. Effects of
chronic hypoxia, normoxia and hype-
roxia on larval development in the beetle
Tenebrio molitor. J. Insect Physiol.
42:991?96
198. Temple GF, Metcalfe J. 1970. The effects
of increased incubator oxygen tension on
capillary development in the chick
chorioallantois. Condor 90:187?92
199. McGrath JJ. 1971. Acclimation response
of pigeons to simulated high altitude. J.
Appl. Physiol. 31:274?76
200. Pionetti J-M, Bouverot P. 1977. Effects
of acclimation to altitude on oxygen
affinity and organic phosphate concentra-
tions in pigeon blood. Life Sci. 20:1207?
12
201. Metcalfe J, McCutcheon, IE, Francisco
DL, Metzenberg AB, Welch JE. 1981.
Oxygen availability and growth of the
chick embryo. Respir. Physiol. 46:81?88
202. Williams JB, Swift K. 1988. Oxygen
consumption and growth of Northern
OXYGEN AND ANIMAL FLIGHT EVOLUTION 155
Bobwhite embryos under normoxic and
hyperoxic conditions. Condor 90:187?92
203. Bigard AX, Brunet A, Guezennec C-Y,
Monod H. 1991. Effects of chronic hy-
poxia and endurance training on muscle
capillarity in rats. Pflu?gers Arch. 419:
225?29
204. Hoppeler H, Desplanches D. 1992. Mus-
cle structural modifications in hypoxia.
Int. J. Sports Med. 13(Suppl. 1):S166?68
205. Weber KE. 1996. Large genetic change
at small fitness cost in large populations
of Drosophila melanogaster selected for
wind tunnel flight: rethinking fitness sur-
faces. Genetics 144:205?13
206. Galun R, Fraenkel G. 1961. The effect of
low atmospheric pressure on adult Aedes
aegyptii and on housefly pupae. J. Insect
Physiol. 7:161?76
207. Dudley R. 1995. Extraordinary flight per-
formance of orchid bees (Apidae: Eug-
lossini) hovering in heliox (80% He/20%
O2). J. Exp. Biol. 198:1065?70
208. Chai P, Dudley R. 1995. Limits to ver-
tebrate locomotor energetics suggested
by hummingbirds hovering in heliox.
Nature 377:722?25
209. Chai P, Dudley R. 1996. Limits to flight
energetics of hummingbirds hovering in
hypodense and hypoxic gas mixtures. J.
Exp. Biol. 199:2285?95
210. Stalker HD, Carson HL. 1948. An alti-
tudinal transect of Drosophila robusta
Sturtevant. Evolution 2:295?305
211. Hamilton TH. 1961. The adaptive signif-
icances of intraspecific trends of varia-
tion in wing length and body size among
bird species. Evolution 15:180?95
212. Feinsinger P, Colwell RK, Terborgh J,
Chaplin SB. 1979. Elevation and the
morphology, flight energetics, and for-
aging ecology of tropical hummingbirds.
Am. Nat. 113:481?97
213. Epting RJ. 1980. Functional dependence
of the power for hovering on wing disc
loading in hummingbirds. Physiol. Zool.
53:347?57